The highest single-year Budget allocation of Rs 48,000 crore for the financial year 2017-18 has removed all apprehensions about the longevity of the MGNREGA scheme, after Prime Minister Narendra Modi called it a “living monument of UPA’s failure” in a Parliament speech on February 27, 2015. Until FY17, the government has allotted a total sum of Rs 3,36,699 crore for the MGNREGA scheme. Given the quantum of the funds, it is natural to assess how efficiently states have utilised those funds and generated the scheme outcomes (rural employment indicators). On these counts, states’ performance seems to vary significantly since the extension of the scheme to the entire country in 2008.
The success of MGNREGA depends on the performance of individual states. What can be a reasonable standard for assessing how well the states are doing? One intuitive way can be to see how well individual states have done in utilising their allotted funds under the scheme to generate the given scheme outcomes. Another equivalent way could be to see how well individual states have produced the scheme outcomes corresponding to the allotted funds.
With no universal benchmark available, individual states’ performance can be evaluated in comparison to the relatively best-performing states in the group in a given year, say, 2013-14. Using the scheme data available for 2013-14, relative performance evaluation for the states has been made using input resources (expenditure on wages, expenditure on material, administrative expenditure) and outputs (number of households provided employment, total person-days of employment, SC & ST person-days of employment, women person-days of employment, number of households that availed 100 days of employment).
You may also like to watch:
Assuming returns to scale constant, the evaluation reveals that 10 states (Arunachal Pradesh, Goa, Kerala, Meghalaya, Mizoram, Nagaland, Punjab, Rajasthan, Tamil Nadu and Tripura) and the Union territory of Puducherry are 100% efficient (in relative terms)—i.e. no other state can outperform these when it comes to the output production from their input resources (funds) utilised. This means that data do not suggest any scope for performance improvement and hence resource utilisation and employment generation is at the optimum level for these states. The remaining 18 states and two Union territories are inefficient. The next seven states with favourable performance (with efficiency of 80% or more) are Andhra Pradesh, Chhattisgarh, Gujarat, Jharkhand, Manipur, Odisha and West Bengal.
It is worthwhile to mention that most of the efficient states are relatively small in size and have a relatively high literacy rate, with the possible exception of Rajasthan.
Therefore, it can be concluded that social awareness about the scheme among the masses plays an important role in the performance. However, the reasoning of small size and relatively high literacy rate does not fully explain the 100% efficiency of Rajasthan and Tamil Nadu. Their good performance is because this evaluation is based on the maximum production of outputs, and these two states generate maximum possible rural employment with the given input resources. The average efficiency of 84.37% shows that states need to increase their employment indicators at least by 15.63% to perform at par with the best performers. This implies, on average, an individual state can increase all its employment figures at least by 15.63% without requiring any additional expenditure. The biggest state Uttar Pradesh, for instance, is only 79.47% efficient. Among the inefficient, the worst-performing state of Karnataka is only 51.96% efficient; thus Karnataka requires almost doubling its employment figures to bring its performance at par with the efficient states.
It is interesting to observe the performance of Jammu and Kashmir and Uttarakhand. Their lower efficiency scores—61.46% and 62.82%, respectively—seem to contradict the earlier argument of relatively small size and decent literacy rate. Close to the bottom performance of these two states can be explained by their topography and other external factors not taken into account in the evaluation methodology. These are primarily hilly states, and it is difficult to continuously provide employment for the rural poor on works related to drilling wells for irrigation, ponds construction, rural roads construction, etc, as compared to the states in the plains.
It is noteworthy to mention that if all inefficient states perform at par with the best practice, then huge gains, in the form of funds saving or extra employment generation, are possible. The analysis shows that, on an average, 17.89% of total expenditure (R4,680 crore) could have been saved in FY14 alone. Most inefficient states perform poorly when it comes to the participation of women and disadvantaged sections (SC & ST) in the scheme. To catch up with the performance of the best, the inefficient states need to enhance women participation by 133%.
As scale economies play an important role in the success of any social welfare scheme, it is important to see the impact of scale on the performance of individual states. Here, the assumption of variable returns to scale reflects a combination of true managerial efficiency as well as the effects of uncontrollable constraints imposed by external conditions. The average scale efficiency is 92.99%. This means an individual state needs to alter the scale of the scheme by about 7%. In other words, an individual state’s actual production scale diverges by 7% from the most productive scale size. Of the 28 states and three Union territories evaluated, only 10 states and one Union territory have optimum scale or maximum productive scale size.
Only Andaman & Nicobar Islands, Lakshadweep and Sikkim operate at increasing returns to scale or sub-optimum scale. For them, scale efficiency ranges from 74.92% to 98.64%, so operate on a scale that is too small to be efficient. One possible reason for their sub-optimal scale is that they are very small by population and area, and as a result could not provide ample employment opportunities to their population due to the lack of available work.
The author is professor, Operations Management Group, Indian Institute of Management Calcutta